By now it is probably safe to assume that robots are considered to be extremely reliable and that most of us look forward to a future when every one of us will have our own steel-based personal assistant to walk around with and do everything for us.
They’ve proven they can build houses and cars and vacuum our homes, so the possibilities must be endless when talking about a robot’s ability. Since their processes are automated, this means we trust them to function accurately, which implies a high degree of trust from our part.
A group of scientists from Georgia Tech decided to conduct a series of experiments to see just how high our trust in robots really is.
Their study focused on 42 volunteers who were told to follow a “guidance robot” into a conference room. In order for the experiments to be successful, the participants were not informed about the true purpose of the test.
During the experiment, the robot circled round a few times, then even led people to the wrong location and at some point it even stopped functioning. One of the researchers then told the participants that the robot had broken down.
This should’ve typically hindered the participants’ trust in the robot, but as the experiment demonstrated, that was not the case.
At one point, while the subjects were in the conference room, the researchers filled the corridor with artificial smoke and set off the alarm. As they started exiting the room, they all followed the robot, which was supposed to guide them towards the exit.
That did not happen, however. The robot actually steered them in the opposite direction and all the participants followed the wrong route. Even though the emergency exists were suggestively marked and even though the robot had malfunctioned before, the participants still placed a great amount of trust in the robot’s skills.
According to Paul Robinette, one of the research engineers involved in this study, these results were completely unexpected. Their findings suggest that as long as a robot informs someone they can perform a particular task, people are willing to believe it and trust its abilities, even if it malfunctions.
Nevertheless, seeing how most people trust their GPS systems so much, it isn’t surprising to see this confidence transferred into another technology, as well.
These experiments are the first ones to focus on human-robot trust in a real emergency situation and will be presented at the International Conference on Human-Robot Interaction, which will take place next week in New Zealand.
Image Source: Engadget